How much noise floor and other things matter in oscilloscope usability

[Fungus]

OTOH you could even go down to a 500 Euro oscilloscope like the Siglent SDS1204X-E. It's probably all you need and you'd have enough money left over to create an awesome lab.

Yes I know, this is one of my doubts. I will check some review of that scope but in the mean time what do you think about its bigger brother sds2000x-plus?

Me? For that money I'd start with the little brother plus 1000 Euros of other toys(!). When you know more about oscilloscopes and their limitations you can decide if you want to sell it and buy the bigger one.

[Fungus]

4) 100mv/div, probe, all the same

Looks flat enough to me.  Nothing to be unhappy about.

What about 1v/div with 10x probe?

[Fungus]

Also try averaging mode:



(image from section 4-5 of the manual)

Averaging can be used with any periodic waveform, ie. almost all waveforms where noise might be an issue, eg. measuring ripple.

[Fiorenzo]

With 1mv/div It gives about 160mV RMS and 1,2mV peak to peak. Whatching other scopes they seem to have about 4 time less noise.
Am I wrong?

About me I am not rich, but I would like to buy a quite good oscilloscope.

I dislike to spend a lot of money ofcourse, but in my experience with quality instruments you get the best results from what you are doing so maybe a big effort now could be a big help tomorrow, or maybe It Is better that now I spend a little amount of money and wait for more evolute oscilloscopes in the future....

If the 1000 series is good for me I take It, if It Is better the Rigol or the 2000 series i will go for It but at the moment I am unsure on which way to take.

You speack about other usefull instruments, what could I buy?
I have the solder station and some fixed voltage psu and one psu I have made by my self with 4 indipendent regulable output but not very precise....

[bdunham7]

Here I am, I did the test you asked for, please let me know whath do you think about It.
Thank you very kind!

So just as a quick apples-to-apples comparison with your first example, I set up an SDS1104X-E with the same no-probe, 20MHz BW, 5us/div and 1mv/div.  Then I connected a 1mV_rms 100kHz signal so you can see the suitability of the scope for viewing such a signal.  The noise makes it a bit jumpy and jittery because it affects the trigger point.  I have some low-frequency noise that isn't from the scope but is from other stuff that I have around that I can't turn off right now, so I did both a live snapshot and a stopped picture, which is effectively a single-shot.  As you can see, the scope is perfectly usable on 1mV signals and your Rigol, which is over 3X as noisy, might have a tough time with the same signal.

I also took a snapshot with the input set to GND, which shows you the ADC noise, or so I think.  The LSB at this setting is 40uV, so the peak-to-peak noise is 1 bit. 

I can't tell you what scope works best for you, but if you work at all with small signals, the front-ends of the entry level Siglents is definitely the best for the buck.  Features like the ERES acquisition mode and the 10-bit mode of the SDS2000X+ series do make them better, but you shouldn't generally count on multi-capture averaging even though it can look very good, as this can lead to errors.

[Fungus]

With 1mv/div It gives about 160mV RMS and 1,2mV peak to peak. Whatching other scopes they seem to have about 4 time less noise.
Am I wrong?

It's not a question of right/wrong, it's a question of whether it makes a difference.

There's a whole thread here saying "not as much as you think".

About me I am not rich, but I would like to buy a quite good oscilloscope.

Of course.

If the 1000 series is good for me I take It, if It Is better the Rigol or the 2000 series i will go for It but at the moment I am unsure on which way to take.

It's impossible to define "better" when there's almost 50% price difference. 

Is a car with leather seats "better" than a car which doesn't have them? You can argue that it is but most people still do the shopping with ordinary seats and can fit just as much in the car.

You speack about other usefull instruments, what could I buy?
I have the solder station and some fixed voltage psu and one psu I have made by my self with 4 indipendent regulable output but not very precise....

What multimeters do you own? Do you have space for a bench multimeter? A nice bench multimeter will be far more useful than a Siglent 2000 series vs. a Siglent 1000 series. A good handheld meter is also good (and both together is even better).

A good power supply is also more useful. Not so much for providing power but for setting current limits and experimenting with "will this run at XX volts".

You can also afford to get the logic analyzer add-on for your Siglent 1000 series. That will be great for your digital work.

What solder station have you got? A 100 Euro solder station is much better than a 20 Euro station.

Hot air gun for desoldering?

Thermal camera? More useful than you might think...

[tautech]

You can also afford to get the logic analyzer add-on for your Siglent 1000 series. That will be great for your digital work.

Don't, it's an abomination get the SDS2104X+ and SPL2016 instead. Totally different and better class of MSO probe.

[Fungus]

Don't, it's an abomination

Tautech says something made by Siglent is an abomination? Avoid like the plague. 

(I haven't actually used one, I sorta assumed it would work for 300 Euros given that the rest of the 'scope is OK)

get the SDS2104X+ and SPL2016 instead. Totally different and better class of MSO probe.

And ... totally blows the budget. That's the tautech we know and love. 

Edir: I'm only kidding, OP did mention "2000 Euros" somewhere and it's not quite 2000...

[Fiorenzo]

Hahahaha you all good guys!
I like you

[Fungus]

Hahahaha you all good guys!
I like you

Have you tried averaging mode yet? 

[Fiorenzo]

So, I did some test, just to learn and experiment.

Same crappy psu, two probes ON in DC mode, one disconnected and the pink one probe on the psu output.
First photo 10x attenuation, second photo 1x.

The ripple of the psu is about 150mv, now the front end noise seem to make a good part of the noise saw on the ripple wave, high resolution or avarage get rid of a big part of this noise. If I did not be aware of the noisy front end i could think that the noise in the ripple was from the psu itself.
Second if i work with the probe in 1x mode the ripple Is now very evident.
Is this the right way to check It? I mean in 1x mode.
Should i use AC or DC mode?
I have read that AC mode get rid of the DC bias so good to work in this way...

[Fiorenzo]

Hahahaha you all good guys!
I like you

Have you tried averaging mode yet? 

Yes many times, and also hi-res mode.
I didn't used It in the photos because i wanted to show the real front end noise.

Actually i really like this rigol, Is a powerfull scope, but i have never put my hands on the siglents

[Fungus]

If I did not be aware of the noisy front end i could think that the noise in the ripple was from the psu itself.

Ripple can only be at the switching frequency of the PSU, never in the MHz range.

Second if i work with the probe in 1x mode the ripple Is now very evident.
Is this the right way to check It? I mean in 1x mode.

Yes. 1x mode with 20MHz bandwidth limiter and maximum averaging.

Should i use AC or DC mode?
I have read that AC mode get rid of the DC bias so good to work in this way...

AC will center the ripple waveform around zero so you measure RMS, etc.

(ie. it's good)

[Fiorenzo]

Here I am, I did the test you asked for, please let me know whath do you think about It.
Thank you very kind!

So just as a quick apples-to-apples comparison with your first example, I set up an SDS1104X-E with the same no-probe, 20MHz BW, 5us/div and 1mv/div.  Then I connected a 1mV_rms 100kHz signal so you can see the suitability of the scope for viewing such a signal.  The noise makes it a bit jumpy and jittery because it affects the trigger point.  I have some low-frequency noise that isn't from the scope but is from other stuff that I have around that I can't turn off right now, so I did both a live snapshot and a stopped picture, which is effectively a single-shot.  As you can see, the scope is perfectly usable on 1mV signals and your Rigol, which is over 3X as noisy, might have a tough time with the same signal.

I also took a snapshot with the input set to GND, which shows you the ADC noise, or so I think.  The LSB at this setting is 40uV, so the peak-to-peak noise is 1 bit. 

I can't tell you what scope works best for you, but if you work at all with small signals, the front-ends of the entry level Siglents is definitely the best for the buck.  Features like the ERES acquisition mode and the 10-bit mode of the SDS2000X+ series do make them better, but you shouldn't generally count on multi-capture averaging even though it can look very good, as this can lead to errors.

Ahhahaah bdunham7 with the same settings i get almost 5 times your noise. The Rigol really like noise....

[Fiorenzo]

If I did not be aware of the noisy front end i could think that the noise in the ripple was from the psu itself.

Ripple can only be at the switching frequency of the PSU, never in the MHz range.

Second if i work with the probe in 1x mode the ripple Is now very evident.
Is this the right way to check It? I mean in 1x mode.

Yes. 1x mode with 20MHz bandwidth limiter and averaging mode enabled.

Should i use AC or DC mode?
I have read that AC mode get rid of the DC bias so good to work in this way...

AC will center the ripple waveform around zero so you measure RMS, etc.

(ie. it's good)

Thank you fungus you are helping me very well.

And also merry Christmas in advance to everyone!!

[tautech]

Don't, it's an abomination

Tautech says something made by Siglent is an abomination? Avoid like the plague. 

(I haven't actually used one, I sorta assumed it would work for 300 Euros given that the rest of the 'scope is OK)

It works fine but it becomes another box on the bench instead of being inbuilt which would be difficult with the very compact SDS****X-E series.

get the SDS2104X+ and SPL2016 instead. Totally different and better class of MSO probe.

And ... totally blows the budget. That's the tautech we know and love. 

Edir: I'm only kidding, OP did mention "2000 Euros" somewhere and it's not quite 2000...

SPL2016 with a new scope are on special ATM until at least years end for the unbeatable price of $219 with MSO and FG licensing. That's one hell of a deal unless you want to DIY one with the info from that thread.

Christmas midday dinner calls.......

[Fiorenzo]

Here I am, I did the test you asked for, please let me know whath do you think about It.
Thank you very kind!

So just as a quick apples-to-apples comparison with your first example, I set up an SDS1104X-E with the same no-probe, 20MHz BW, 5us/div and 1mv/div.  Then I connected a 1mV_rms 100kHz signal so you can see the suitability of the scope for viewing such a signal.  The noise makes it a bit jumpy and jittery because it affects the trigger point.  I have some low-frequency noise that isn't from the scope but is from other stuff that I have around that I can't turn off right now, so I did both a live snapshot and a stopped picture, which is effectively a single-shot.  As you can see, the scope is perfectly usable on 1mV signals and your Rigol, which is over 3X as noisy, might have a tough time with the same signal.

I also took a snapshot with the input set to GND, which shows you the ADC noise, or so I think.  The LSB at this setting is 40uV, so the peak-to-peak noise is 1 bit. 

I can't tell you what scope works best for you, but if you work at all with small signals, the front-ends of the entry level Siglents is definitely the best for the buck.  Features like the ERES acquisition mode and the 10-bit mode of the SDS2000X+ series do make them better, but you shouldn't generally count on multi-capture averaging even though it can look very good, as this can lead to errors.

What circuits works with such a low intensity signals?
Sensors for example?
I could think about maybe a shunt, they gives very low output when measuring currents, i use many times shunts in my circuits....

Speaking instead about a signal amplifier, as suggested by some one in this thread, It was told that they are a better way to check low intensity signals than giving faith on a low noise front end.

Is this true all times?

[Fungus]

AC will center the ripple waveform around zero so you measure RMS, etc.

...which is of course the wrong way to do it, and was the big error in the now-infamous "Rigol vs. Siglent noise" video.

Using AC mode to calculate RMS assumes the 'scope is perfectly calibrated, ie. with zero offset voltage.

The correct way to measure RMS is with "Std.Dev", which removes the offset voltage mathematically in both AC and DC modes.



You should still use AC mode though because it centers the signal and allows you to crank up the vertical scale without going off screen.

[Fungus]

Have you tried averaging mode yet?

Yes many times, and also hi-res mode.
I didn't used It in the photos because i wanted to show the real front end noise.

Averaging mode isn't cheating! It's there for a reason... 

(And even Siglents have it)

[Fiorenzo]

Watch this photo attached.

If i don't ground the probe tip the oscilloscope shows a 50hz sinewave. I have also tried turning off every appliance connected near the scope and also the lights but the sine wave remain unchanged.

[nctnico]

I could think about maybe a shunt, they gives very low output when measuring currents, i use many times shunts in my circuits....

Speaking instead about a signal amplifier, as suggested by some one in this thread, It was told that they are a better way to check low intensity signals than giving faith on a low noise front end.

For such purposes you better use a pre-amplifier with a differential input. You can easely DIY these using an instrumentation amplifier chip or a specific current sensing amplifier chip. From there feed the amplified signal into your oscilloscope.

[Fiorenzo]

Don't, it's an abomination

Tautech says something made by Siglent is an abomination? Avoid like the plague. 

(I haven't actually used one, I sorta assumed it would work for 300 Euros given that the rest of the 'scope is OK)

It works fine but it becomes another box on the bench instead of being inbuilt which would be difficult with the very compact SDS****X-E series.

get the SDS2104X+ and SPL2016 instead. Totally different and better class of MSO probe.

And ... totally blows the budget. That's the tautech we know and love. 

Edir: I'm only kidding, OP did mention "2000 Euros" somewhere and it's not quite 2000...

SPL2016 with a new scope are on special ATM until at least years end for the unbeatable price of $219 with MSO and FG licensing. That's one hell of a deal unless you want to DIY one with the info from that thread.

Christmas midday dinner calls.......

Yes i really need a complete mso now because i am designing an e-bike controller from scratch and also working with the display circuitry and program that i am also designing from the bottom. I did similar things in the past, but without an oscilloscope and had to cope with the problems encountered only with the help of the theory and literature due to the fact that i couldn't check my circuits... But i got to the end fine everytime. Such a satisfaction...
Sorry for my bad english...

The offer from siglent is very appealing and i don not have the time now to make the probes by my self.
But i could instead buy a cheaper scope and an external LA to connect on the computer, maybe this  is the best way to proceed at the moment
Ummm... But i am not sure what way i will take hahaha....

Actually i saw that the AWG of the siglent and rigol are limited and an external signal generator could be better.... But i like that the Rigol has two outputs.
I should try a siglent but Amazon don't sell It here in Italy, only the lower models....

[Fiorenzo]

I could think about maybe a shunt, they gives very low output when measuring currents, i use many times shunts in my circuits....

Speaking instead about a signal amplifier, as suggested by some one in this thread, It was told that they are a better way to check low intensity signals than giving faith on a low noise front end.

For such purposes you better use a pre-amplifier with a differential input. You can easely DIY these using an instrumentation amplifier chip or a specific current sensing amplifier chip. From there feed the amplified signal into your oscilloscope.

Thank you i will do It

[bdunham7]

What circuits works with such a low intensity signals?
Sensors for example?
I could think about maybe a shunt, they gives very low output when measuring currents, i use many times shunts in my circuits....

Speaking instead about a signal amplifier, as suggested by some one in this thread, It was told that they are a better way to check low intensity signals than giving faith on a low noise front end.

Yes--sensors, audio circuit, measuring circuits, even RF circuits.  And keep in mind as I said before that you may often want to use a 10X or even 100X probe or custom voltage divider to avoid loading the circuit you are testing and that will magnify the problem. 

Here's the scope with a 10X probe connected to a 50-ohm terminated signal generator with no signal running, no signal single-shot and then the same 1mV signal as before.  Note that I had to use the single-shot to capture the signal as it is so low that the scope won't trigger on it, however it is still clearly visible above the noise. 

As far as preamplifiers, yes they have a place but they are usually fairly specific and limited, as opposed to being general purpose.  For example they typically have a much more limited input voltage range or bandwidth, or both.

[Fungus]

Fiorenzo,

Can you post a screenshot of your PSU ripple with 20MHz limiter and maximum averaging?